I used the open air test method to test the 14-N's wireless performance. As noted earlier, I used an SMCWCB-N2 Cardbus adapter, which was inserted into a Fujitsu P7120 Lifebook (1.2 GHz Intel Pentium M, 504 MB) notebook running WinXP Pro SP2 with all the latest updates. I used the latest Win XP SP2 220.127.116.11 driver and SMC 18.104.22.168 client utility during testing. The router had the latest 22.214.171.124 firmware and I left all factory default settings in place, except to force channel 1.
Figure 9 shows a composite of up and downlink throughput tests made at the six test locations in two wireless modes: 2.4 GHz, 20 MHz channel; and 2.4 GHz, 40 MHz channel. Each column represents the average throughput from a one minute IxChariot test.
You can see best-case (Location A) throughput is 76 Mbps running uplink with a 40 MHz channel. In general, the 40 MHz bandwidth mode produces higher throughput than the default, "legacy" (802.11b/g) friendly mode with stronger signal levels. But throughput gain from channel bonding fades as signal levels diminish.
Figure 9: SMCWGBR14-N wireless throughput mode comparison
5 GHz gear usually can't successfully pass traffic in my most difficult test Locations E and F. But it's unusual for this to happen with 2.4 GHz draft 11n products. But that's what I found with the combination of 14-N and SMCWCB-N2 client. In Locations E and F, the client was able to see the 14-N, but could not successfully pass traffic. Figure 10 shows that I even had problems getting a complete test run at Location D, and even in 20 MHz bandwidth mode!
Figure 10: Wireless throughput - 2.4 GHz, 20 MHz channel, uplink
Throughput stability is nothing to write home about, either, as Figure 11 illustrates. The high variation has the effect of reducing the average throughput readings, which is why the average results are so low for a three antenna product.
Figure 11: Wireless throughput - 2.4 GHz, 20 MHz channel, downlink
Wireless Security Performance
The combination of 14-N, SMCWCB-N2 notebook adapter and SMC's client utility did not do well in wireless security testing. Using this combination, I could associate in WPA/TKIP and WPA2/AES modes, but not successfully lease an IP address.
But when I fired up Windows' Wireless Zero Configuration (WZC) utility, the card was able to both connect and successfully lease an IP address. So I performed wireless security tests in WEP 128, WPA/TKIP and WPA2/AES using WZC, the results of which are summarized in Figure 12 and show the typical knockdown of throughput in WEP and WPA/TKIP modes.
Figure 12: Wireless security tests - 20 MHz mode, downlink
I had to run the tests a few times because I found that the router would automatically set the 802.11n mode to Mixed 802.11g and 802.11b when I enabled WEP security. But it did not change it back to Mixed 802.11n, 802.11g and 802.11b when I selected WPA or WPA2 modes. Even then, I got a little confused when I saw higher than 54 Mbps link rates when connected in WPA2/TKIP, but still 11g rate throughput.
Both the SMC client and the 14-N support Wi-Fi Protected Setup (WPS). And when I ran a push-button WPS session, everything appeared to complete successfully with a WPA TKIP connection reported. But, as with the manual connection, the client wasn't able to lease an IP address through the encrypted connection. I wasn't able to test WPS using WZC, since WZC doesn't support WPS, even in Vista.